Research map / 3D vision / 3D generation
TextSLAM: Visual SLAM With Semantic Planar Text Features
Boying Li, Danping Zou, Yuan Huang, Xinghan Niu, Ling Pei, Wenxian Yu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add text features to point-only SLAM (ablation) | TextSLAM · Indoor tests Indoor_01 | APE (m): 0.086 -> 0.067 | Improved |
| add text features to point-only SLAM (ablation) | TextSLAM · Indoor tests Indoor_05 | APE (m): 0.072 -> 0.058 | Improved |
| add text features to point-only SLAM (ablation) | TextSLAM · Indoor tests Indoor_07 | APE (m): 0.045 -> 0.032 | Improved |
| add text features to point-only SLAM | TextSLAM · Indoor tests Indoor_10 | APE (m): 0.076 vs 0.094 point-only; DSO 0.071 best | Hurt |
| add text features under severe motion blur (ablation) | TextSLAM · Rapid motion tests | APE (m): point-only fails on all three sequences; TextSLAM 0.060, 0.020, 0.058 | Improved |
| track text as a whole object instead of ORB or DSO points | TextSLAM · Rapid motion tests Rapid_03 | APE (m): ORB-SLAM 0.085, DSO 0.113 -> 0.058 | Improved |
| add semantic text-based loop closing | TextSLAM · Small indoor loop test AIndoorLoop_02 | APE (m): 0.075 -> 0.018 | Improved |
| add semantic text-based loop closing | TextSLAM · Large indoor loop test LIndoorLoop_03 | APE (m): 2.102 -> 0.192 | Improved |
| add semantic text-based loop closing | TextSLAM · Outdoor day tests Outdoor_1 | APE (m): 1.159 -> 0.688 | Improved |
| add text-based loop closing when viewpoint change is small | TextSLAM · Outdoor day tests Outdoor_4 | APE (m): 0.116 -> 1.624 | Hurt |
| estimate text plane as a whole instead of RANSAC plane fitting to ORB-SLAM/DSO points | TextSLAM · 3D text map quality | angular and distance error | Improved |
| use 3D text map matching instead of NetVLAD-based localization | TextSLAM · Day-night localization (987 day, 145 night keyframes) | precision/recall | Improved |
Closest papers by meaning
- SNI-SLAM++: Tightly-Coupled Semantic Neural Implicit SLAM (TPAMI 2025 · similarity 0.86)
- SNI-SLAM: Semantic Neural Implicit SLAM (CVPR 2024 · similarity 0.85)
- G^3-LQ: Marrying Hyperbolic Alignment with Explicit Semantic-Geometric Modeling for 3D Visual Grounding (CVPR 2024 · similarity 0.83)
- More Text, Less Point: Towards 3D Data-Efficient Point-Language Understanding (AAAI 2025 · similarity 0.83)
- MSTAR: Box-free Multi-query Scene Text Retrieval with Attention Recycling (NeurIPS 2025 · similarity 0.83)
- FineLIP: Extending CLIP's Reach via Fine-Grained Alignment with Longer Text Inputs (CVPR 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
